I am testing a 10x fastq dataset , but cellranger count complains "An extremely low rate of correct barcodes was observed for all the candidate chemistry choices for the input", I have tried all the chemistry recommendation codes on the 10x genomics website, and I still get the same result, how can I solve this problem, thanks.

enter image description here enter image description here

  • $\begingroup$ It is difficult to provide an answer when the question does not include what the input is, what command was run, ... I can only suggest to use the --chemistry=auto option of cellranger count. $\endgroup$
    – haci
    Dec 12, 2021 at 14:03
  • $\begingroup$ What does the quality look like on those reads? If this is 10XGenomics library, they should be able to help your troubleshoot. $\endgroup$
    – swbarnes2
    Dec 12, 2021 at 22:44
  • $\begingroup$ Could you please give a few more details about the nature of this project? Are you sure this is a single-cell dataset (I'm aware that 10X do other non-single-cell things as well)? Showing the first few lines of header from the fastq files can be very helpful to work out the structure of the experiment. $\endgroup$
    – gringer
    Dec 12, 2021 at 23:41
  • $\begingroup$ how were your fastqs generated? and how are they named? $\endgroup$ Dec 13, 2021 at 21:01
  • $\begingroup$ Thanks @ciwei for providing your answer. Could you kindly check the "accept" box for your answer? This does help site statistics $\endgroup$
    – M__
    Sep 2, 2023 at 13:34

2 Answers 2


I have solve this problem, cause the combined barcode is differernt from the common 10x barcode. when mapping the genome, need to add-- chemistry ARC-v1 when cellranger count.

  • 1
    $\begingroup$ Excuse me, how do you determine the required chemistry? $\endgroup$
    – JY_sea
    Apr 6, 2022 at 6:10
  • $\begingroup$ ARC-v1 means this was done using a multiomic kit. From cellranger count --help "To analyze the GEX portion of multiome data, chemistry must be set to 'ARC-v1'; 'ARC-v1' chemistry cannot be autodetected" $\endgroup$
    – nico
    Nov 1, 2023 at 16:44

Answer from @maximilian-press, converted from comments:

I can't comment on 10X data specifically, however, I suggest writing a simple script or command that counts the frequency of each barcode, and reports back the N (10? 20? 100?) most frequent ones, which will allow you to investigate in more detail. This is what I have done in the case of suspected bardode swaps or reverse complement issues.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.